An underdiscussed aspect of covid policy was the speed with which government supported people financially and temporarily solved social problems. Furlough is widely seen as a success. Rishi Sunak built his reputation on it. But what’s really remarkable about it is how quickly and seamlessly it was rolled out. It was easy for businesses to use, and beneficiaries barely had to do anything at all. Likewise the additional £20 a week for those using Universal Credit was paid out without complications to all who needed it, and, briefly, reduced poverty. Rough sleeping was halved within a few weeks once it became a threat to the rest of society, though it’s now gone back up again.
This should leave us with a big question. If it was so easy then, when government was hardly operating at maximum effectiveness, why is so hard the rest of the time? How much of the misery that is inflicted on those who need government support in normal times due to poor administration and how much is deliberate?
Political scientists have taken an increasing interest in these questions in recent years and have developed a language for talking about what is often a hidden world of policy decisions. In this literature, deliberate attempts to create barriers to accessing services are known as “ordeals”. An example would be the decision to prevent anyone seeking asylum from working, so forcing them to live on a pittance, in the hope that it will put people off attempting to come to the UK, even though doing so adds considerable cost to the state. The welfare system is full of ordeals, whether in the form of complex assessments needed to access services, or the threat of sanctions, for those receiving benefits, if various conditions aren’t met.
Ordeals have become an increasingly common tool for policy-makers and this leads to two concerns. First is the lack of transparency and accountability. Ordeals often don’t require any legislation or debate. They just involve making life more difficult for particular groups in the hope this reduces demand. The fact that, for instance, new applicants for universal credit have to wait for five weeks before getting support, or apply for a loan, is not widely known by those who have no interactions with the system. Sometimes the government go out of their way to pretend ordeals don’t exist – such as when they brief newspapers that asylum seekers are being put up in luxury hotels when the reality is conditions are made deliberately unpleasant for the vast majority. Or, as with voter ID, pretend ordeals have one purpose (stopping non-existent fraud) when they have another (trying to reduce the numbers voting for opposition parties).
The other problem, even if you accept the ethical compromises involved in using ordeals, is that they often don’t work, usually because the policy designer has misunderstood the motivations of those in need. And indeed they can often make policy problems worse and substantially increase costs for the taxpayer. The rest of this post focuses on why ordeals rarely work, and why, nevertheless, politicians, policymakers, and to some degree the wider public, keep thinking they will.
Why ordeals fail
In a recent report – based on deep personal experience of helping people through the benefits system – Kayley Hignell, the Director of Policy at Citizens Advice, makes a critical point. It is one that is not complicated and yet is widely ignored in policy-making:
“If you want people to focus on improving their income from work, it’s arguably more helpful to take actions that reduce the demands on their cognitive bandwidth, and free up more space in their minds to think about other things….Some things that use up a person’s cognitive bandwidth are actually caused by other parts of the benefit journey. For example, the challenges of figuring out levels of childcare support and payment cycles that are out of sync with wages and bills. Things like this cause repeated budgeting nightmares for people. There are hundreds of policy trade-offs like this, hidden in the detailed design and processes of claiming and maintaining benefits. Currently, we don’t factor in the downside of overloading people's cognitive bandwidth against trying to make a system easy to deliver.”
There are mountains of research about the costs of this type of cognitive overload, much of it was collected a decade ago by Sendhil Mullainathan and Eldar Shafir in their book “Scarcity”. Not having enough money by itself reduces bandwidth because you’re thinking about that absence a lot which reduces time for good decision making. As they put it “time for reflection is a luxury good”. While the psychological experiments that prove the point may be recent this idea is hardly new - it’s a major theme of Orwell’s “Down and Out in London and Paris”. It will be familiar to anyone who has experienced scarcity themselves or spends time with those who do.
In the US the Biden Administration has acted in accordance with the research by asking federal agencies to reduce burdens, following another book “Sludge” written by Cass Sunstein, an adviser to Obama and now Biden. This explicitly requests that agencies:
“Describe and discuss sources of psychological costs that certain information collections impose on individuals, such as the cognitive load, discomfort, stress, or anxiety a respondent may experience as a result of attempting to comply with a specific aspect of an information collection.”
Sunstein is now advising the Administration on reducing regulatory burdens in the immigration system. Most benefits in the US, though, are distributed by states and many red states are actively looking to increase ordeals for those trying to access financial and health benefits. For instance Arkansas attempted to introduce work requirements for Medicaid leading to significant loss of access to healthcare until it was stopped by the courts.
In the UK our government has seemed largely uninterested in the costs of ordeals and burdens. Hignell notes, as just one example, the decision to force the several million households migrating from legacy benefits to universal credit to complete a whole new set of forms, or lose their money, rather than automating the process. But perhaps the best example of how ordeals have been misused over the past few decades is disability benefits.
Since the Thatcher-era deindustrialisation of parts of the country the numbers on disability benefits have been growing. The percentage of the working population in receipt has increased, according to the IFS, from 2% in 1992 to 6% now. This is expensive and successive governments have tried to manage costs by introducing more and more ordeals into the process of getting these benefits. In 1995 the Major government introduced Incapacity Benefit that required claimants to be assessed by a government doctor rather than their own GP. In 2008 Labour replaced this with the Employment Support Allowance (ESA) that split claimants into two categories, the less serious of which saw their additional support phased out in an attempt to encourage people back to work.
This has entirely failed to reduce the numbers or cost. ESA created an enormous amount of fear and anxiety and thus a perverse incentive to get into the “serious” group which comes with greater benefits but no expectation of work. It pushed people further away from the labour market. Meanwhile the cost of the actual assessments is huge. The contracts in 2016 were worth £1.6 billion over three years.
In 2014 the Coalition government decided to exacerbate the problem by changing the other disability benefit – the Disability Living Allowance (DLA) which is for additional living costs and paid regardless of whether you work – to the Personal Independent Payment (PIP) which is also much more regularly assessed.
These changes combined have seriously depleted cognitive bandwidth for disabled people and created an expensive industry around trying to secure support for them. It is notable that most of the increase in successful disability claims in recent years has been due to increasing mental health issues. This is a growing challenge across society but it’s hard to believe the policy ordeals imposed by government haven’t made this worse.
Indeed this has been a repeated finding of every review into the system. The work and pensions select committee has come back to the same point over and over:
“Some [claimants] felt that the progress they were making towards recovery, and then moving back into work, was hampered rather than aided by the anxiety caused in facing the WCA (work capability assessment). While it might be expected that those who believed that they were placed in the wrong group felt aggrieved, we also heard from those who felt that they were placed in the most appropriate group in the end, but who also felt that the process had still been a harrowing and stressful ordeal for them.”
“Claimants told us they had found filling in the forms stressful. We heard they are reticent to share their—often very severe—impairments in day-to-day life, instead focusing on what they are able to do and remaining positive. Filling in the form requires claimants to present a record of all the things they struggle with or are unable to do, which can be highly distressing and damaging to self-esteem.”
In 2022 (in an evidence session):
“To get good employment outcomes out of the system, people need to be willing to take chances. They need to have a space for safe experimentation to see whether things work out and to know that if things do not work out, the benefits system will be there for them. That is the opposite of the system that we have at the moment….The lack of trust means that people are more likely to cling on to the benefits that they have, because they do not understand why they were awarded them and they have no trust that they would get the benefits that they think they need if they were to try working.”
There is a temptation, especially for those on the right, to see this all as special pleading from bleeding heart liberals. But it is confirmed in every piece of evidence we have – including the substantial rise in costs of the disability benefits system. Look at the evidence of sanctions, the primary ordeal through which conditions are placed on people, including disabled people, across the welfare system. Report after report has shown that sanctions not only do not have long-term economic benefits, but actively make things worse by increasing hardship and economic inactivity.
Nor is this finding limited to academic assessments. The National Audit Office have said the DWP hasn’t done enough to show that sanctions have any economic benefits. The DWP’s own evidence almost certainly shows that they don’t, though we can only assume that as they have refused to publish it to the point of intervention by the Information Commissioner. If you care about waste and poor use of public money, as Conservatives are supposed to, then this should bother you. It shouldn’t just be left to professional welfare campaigners to make the point.
Budget contradictions
This takes us to the big policy flaw in last week’s Budget, which somehow managed to simultaneously acknowledge this problem while at the same time making things worse.
The government are, rightly, worried that people are dropping out of the labour market. We have not far off a million more people under-65 not working than pre-pandemic. There is some debate as to the exact cause of this, both increasing ill-health and early retirement seem to be factors, but it is causing serious economic issues, and isn’t happening in other countries. Indeed one of the most Alice in Wonderland things about our hyperbolic immigration debate is that it’s the only thing keeping the labour market afloat at the moment, which is why more professions are being added to the visa list and the OBR is assuming further increases in net migration.
Most of the social policy measures in the budget were designed to get more people working. This in itself is an implicit acknowledgement that increasingly punitive ordeals have not worked. The proposals around disability benefit mostly involve increasing support to help people who have some capacity to work with extra coaches, more occupational health, making jobcentres more accessible and so on. Likewise, on childcare, as well as a big expansion of the system for all working parents, there is an increase in childcare allowances for those on universal credit.
So why I am complaining? Well despite this effective acknowledgement that more support rather than ordeals is the way to help people back into work, both the disability benefit and childcare proposals increase the number of people under threat of sanction. Which is entirely contradictory.
The centrepiece proposal on disability is to scrap work capability assessments and use the PIP assessment to cover both living costs and the costs of not being able to work. This is technically extremely complex because the two assessments currently have a different purpose, and won’t happen before the election, but in principle this could reduce burdens. As the White Paper, published alongside the Budget, says:
“We will use this opportunity to build greater levels of trust between DWP and the people who use our services, by reducing the assessment burden that people currently face. We want to introduce a more tailored approach, to allow work coaches to build a relationship with an individual and determine what, if any, work-related activities an individual can participate in.”
At the moment though, while the process is miserable, if you “pass” your work capability assessment then you get benefits without risk of sanction. Under this new proposed system whether or not to sanction will be entirely at the discretion of the DWP. Which has unsurprisingly, given the very low levels of trust disabled people have in the DWP, created a lot of anxiety and concern. More support is good but significantly widening the group of people at risk of sanction exacerbates all the problems of the current system.
The same is true of the childcare plans. More support for working parents on universal credit has been accompanied by an increase in the risk of being sanctioned for those not working. In the rather weaselly language of such documents the budget announced that:
“The government will also strengthen job support for claimants that are lead carers of young children who currently have no or limited requirements to search for and prepare for work.”
I.e. they will increase conditions on parents (mostly mothers) of pre-school children and sanction them if they don’t meet those conditions. And just in case anyone was in any doubt that the DWP and Treasury were backtracking on their commitment to ordeals the Budget also contained a pledge to:
“Strengthen the way the sanctions regime is applied, by automating parts of the process to improve efficiency and reduce error, and ensuring that Work Coaches have the tools and training to implement sanctions as effectively as possible.”
So we have ended up in a somewhat tortured position where the Budget simultaneously implicitly admitted the current system of ordeals wasn’t working while also creating more ordeals. It offered more support to those not working but cancelled out that support by increasing the threat of sanctions on them, and thus cognitive overload.
Why do Policy-makers keep thinking ordeals will work?
We’re left with the question of why government can’t break free of this belief that ordeals are necessary to achieve their goals. Some would no doubt argue that politicians know full well that sanctions don’t work but are cynically looking to exploit narratives around benefits scroungers and fraud. But I think they really believe they do work, depsite the evidence.
Why? Because for people who are financially secure incentives do largely work as they’re supposed to. None of us are completely rationally beings but look at the tax system – rich people are pretty good at changing their behaviour in response to incentives. Here HMRC do need to keep chasing ever more inventive avoidance schemes, and tax fraud costs us vastly more than benefits fraud.
But for the population using the welfare system the reason they are not working is only very rarely because they don’t want to. On the whole they don’t need to be incentivised but helped. And trying to create incentives through ordeals, whether via admin burdens or threat of sanction, makes it harder for many to engage. The fear that you might lose your benefits because, for instance, you’re late to an appointment due to traffic, is debilitating.
The inability to understand this is an empathy gap. It’s not about sympathy – it’s entirely possible to sympathise with the plight of someone in need of welfare but also think the right approach for their own benefit and the taxpayer is a tough sanctions-based regime. But empathy requires actually understanding what life is like for someone in that position. One clever study, that made use of the fact that in Denmark local politicians get to decide on whether people have do mandatory work assignments to get benefits, found that politicians with personal experience of the welfare system were significantly less likely to support burdens than those who hadn’t had any.
In the UK our Parliament has become more diverse in many ways over the last few decades (in 1996 there were more MPs called John than there were women). But not when it comes to experiences of welfare or poverty. If anything it’s got worse, particularly on the Labour side as politics has professionalised and the trade union movement has become less important in supplying MPs. In 1983 only half of Labour MPs had degrees, now it’s 90%.
It’s a topic covered well in Isabel Hardman’s “Why We Get the Wrong Politicians”. She notes that the MP selection process, in all parties, requires a huge amount of unpaid work and, increasingly, fees to companies who help produce glossy literature and professional-looking videos. Getting more people into Parliament who understand the experience of burdens and ordeals on family members and friends is just one reason why changing the way MPs are picked is important.
Politicians are also responding to public attitudes formed in part by media portrayals of benefits recipients – which hardens the empathy gap. This does seem to be starting to change now though. Public attitudes, which have traditionally been very sceptical of those on benefits have become more supportive in recent years. The British Social Attitudes survey found in 2011 that 62% of people thought benefits were too high and were discouraging people from getting jobs. Just 19% thought they were too low and causing hardship. By 2020 this had flipped to 51% too low and 42% too high. YouGov polling shows 59% of people think it’s difficult for those on benefits to increase their income compared to 21% who think it’s easy. When it comes to disability benefit the public is even more supportive with 49% saying there is too little support and just 7% saying too much.
This is a reaction to benefits genuinely become much less generous. But there has also been a shift away from simplistic media portrayals of poverty. Both ITV and the BBC have, during the current cost of living crisis, done a good job of telling peoples’ stories. Even right wing newspapers who have been happy, in the past, to run endless stories about “scroungers” and benefits fraud, have done so less in recent years. This is partly because some people with experience of scarcity have managed to infiltrate the media world (see this by Catlin Moran in The Times for instance). One would hope it’s also because the evidence the current approach doesn’t work is becoming overwhelming.
As we have seen in this week’s budget though there is still much to do to make the world of ordeals and burdens visible to the wider public. And to get politicians to acknowledge that they rarely achieve their goals while often making things worse and costing the taxpayer a lot of money. There is a constant need to tell the stories of those who need state support the most, to show what it’s like to experience scarcity, to show that creating more fear and anxiety makes it harder for people to help themselves, to keep battering away at the empathy gap.
Not to disagree, but to add - even where ordeals are not consciously seen as beneficial, it is difficult to direct scarce policy-maker time and attention, and money, into reducing burdens and making systems easier to access.
There is always pressure to be able to announce or roll out a new initiative, and the amount of labour involved in building systems that talk effectively to one another and enable automation is often invisible to a team of civil servants designing a policy and almost always completely unimaginable to the politicians directing them. By the time the practical challenges of delivery are realised, everybody is committed to unreasonable timeframes, and rather than step back and fix the access problems the tendency is to push ahead and accept a messy or poorly-joined-up system. And then of course it's much harder to fix later, particularly as waves of technical debt build up across government.
For this reason I find it frustrating that people keep citing the speed of policy-making during Covid as a good thing: the Covid solutions were poorly thought through (of necessity, there was no time), some were incredibly effective and some faceplanted at once, the situation forced people to experiment (which is good) but also pushed us towards short term solutions and a 'pay for it later' mentality which left everything vulnerable to being slashed back in a swing back towards austerity once the immediate crisis subsided.
During Covid we had to use the levers we had, which meant that some groups unexpectedly benefitted and others suffered not because of any intent but because there was no existing lever and no time to build one (such as the newly self-employed, who missed out dramatically on furlough support).
So yes, absolutely, let's be more joined-up and stop letting arbitrary cruelty cost us more in the long run - but to do that effectively we need to really get to grips with the technical and practical investment needed to deliver that smoother, more generous system.
Really excellent to highlight the empathy gap, Sam, something not discussed enough. I’d add it extends to a variety of other issues, and suggests there are bigger, wider problems with our political culture and systems.
An example would be medics going on strike. Now I think people are broadly supportive of, say, junior doctors. But I think a problem in the past has been little real, lived understanding of exactly how stressful the job is, and how bad the compensation is compared to it. We all know being a doctor is stressful but I think it’s hard to appreciate unless you work in healthcare or have a close personal relationship with someone who does. My partner works in healthcare, and even I struggle sometimes to empathise with her because the pressure she’s under is so extreme compared to anything I’m used to.
Something I’ve become uncomfortable with in politics not just in the UK but elsewhere is a tendency to generalise. There’s always this sense that there are ‘real people’ who need to be spoken to. But this is simply not true. We live in diverse, complex societies where people have very different professions, personal backgrounds and social relationships. We have to recognise that, and any political institution has to be able to acknowledge this diversity and represent it.
Maybe we need to find a way to bring actual people from diverse personal and professional backgrounds in, not just as elected officials but in a consultative and deliberative way. An example I think of is in criminal trials, where juries become more empathetic towards defendants after they’ve heard their side of the story compared to just reading media reporting of the trial.